Hyperscalers will spend over $1 trillion next year.

A Wharton economist says that bet requires a 2.7x productivity jump to break even.

Current AI revenue runs $150–200 billion a year, far short of what trillion-dollar capex needs to earn a 15% return.

Jessica Wachter says a shortfall risks missed interest payments and possible bankruptcy.

How each outlet framed it
MIT Technology Review
quantifies hyperscaler productivity requirement (2.7x by 2030) against IT boom growth rates, flags bankruptcy risk

Sources: MIT Technology Review